search

yunzhang813 / FUNNEL-GSEA-R-Package

Zhang et al. 2017. FUNNEL-GSEA: FUNctioNal ELastic-net regression in time-course gene set enrichment analysis. Bioinformatics, 33(13), 1944-1952.
https://doi.org/10.1093/bioinformatics/btx104
6 stars 2 forks source link

data is too sparse #1

Open animesh opened 7 years ago

animesh commented 7 years ago

facing the following error with an input matrix of 4560*39 

...
Not enough points in local window, please increase bandwidth!
...
data is too sparse, retry with larger bandwidths!
...
Not enough points in local window, please increase bandwidth!
Error: the data is too sparse, no suitable bandwidth can be found! Try Gaussian kern instead!
Error: FPCA is aborted because the observed data is too sparse to estimate the covariance function!
Error in matrix(rep(mu, m), ncol = m) : 
  'data' must be of a vector type, was 'NULL'

must note that i have only two time points with 10 and 29 samples representing each time point respectively

yunzhang813 commented 7 years ago

Hi Ani,

You have too few time point to define a functional data. There's not enough information to make inference. For this type of analysis, we usually require time-course data with at least 7 time points. However, we are currently working on the same problem with limited time point (1~2 time points). We will update the repository once we have some established work.

animesh commented 7 years ago

Thanks for the clarification @yunzhang813 :+1: Looking forward to the two time point version then :) Meanwhile what would be your recommendation for performing such analysis, any other tool which i can try?